The penetration of DEET through split-thickness cadaver skin was measured in non-occluded Franz cells placed either in a fume hood or on a laboratory workbench. DEET, dissolved in a small volume of ethanol and spiked with (14)C radiolabel was applied to skin at doses from 0.02 to 11000 microg/cm(2). DEET penetration was greater for cells placed on the workbench, and the percentage of radioactivity penetrated after 72 h increased gradually with dose, for doses up to 680 microg/cm(2). At higher doses, it declined. Percent penetration ranged from 11.5 +/- 3.2% for a dose of 0.021 microg/cm(2) in the fume hood to 71.9 +/- 5.5% for a dose of 260 microg/cm(2) on the workbench. Results were interpreted in terms of a diffusion/evaporation model having three parameters-a solubility value for the chemical in the upper stratum corneum, M(sat); a mass transfer coefficient for evaporation, k(evap); and a characteristic time for diffusion, h(2)/D. The parameters obtained from fitting the model to the data (normalized to the fume hood environment) were M(sat) = 18 microg/cm(2) and k(evap) = 2.6 x 10(-5) cm/h. The value of h(2)/D decreased from 16 h at a DEET dose of 25 microg/cm(2) to 10 h at 1480 microg/cm(2), consistent with an increase in skin permeability of about 1.5-fold over this dose range. This effect was confirmed by means of an additional study in which skin samples pretreated with increasing amounts of unlabeled DEET were washed and redosed with (14)C-benzyl alcohol. A small (1.7-fold), but significant, increase in benzyl alcohol penetration with increasing amount of DEET was obtained. Thus, DEET enhanced its own skin permeation rate as well as that of another compound, but the effect was modest and not likely to be a major concern for compounds coadministered with DEET.